TestsBenchmarks · Knowledge
Vals Public Benefits Bench v1.1
Can a model correctly help people navigate public-benefit (US SNAP) rules and eligibility - policy-text application.
% solvedHigher is betterCounts toward:Weights: Research & analysis ×0.6The test's websiteOfficial page ↗
Measured by independent testersIndependent Reported by the makerVendor-reported
| Thinking levelSetting | |||||||
|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5 | Anthropic | 76.9% | Maxeffort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 2 | Claude Fable 5.1 | Anthropic | 74.9% | Maxeffort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 3 | Claude Opus 5.5 | Anthropic | 70.6% | Maxeffort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 4 | Claude Fable 5 | Anthropic | 70.4% | Maxeffort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 5 | Gemini 4 Argon | Google (Gemini / DeepMind) | 69.8% | Highreasoning_effort=high | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 6 | MiMo-V2.6-Pro | Xiaomi MiMo | 68.9% | Defaultvals id xiaomi/mimo-v2.6-pro | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 7 | Hy4 preview | Tencent Hunyuan | 68.6% | Defaultvals id tencent/hy4-preview | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 8 | GLM-5.3 | Z.ai (Zhipu) | 68.5% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 9 | Muse Spark 1.2 | Meta (Meta Superintelligence Labs, Muse) | 68.5% | Extra highreasoning_effort=xhigh | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 10 | Kimi K3 | Moonshot AI (Kimi) | 68.3% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 11 | Claude Opus 4.8 | Anthropic | 68.1% | Maxeffort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 12 | MiMo-V2.6-Flash | Xiaomi MiMo | 67.6% | Defaultvals id xiaomi/mimo-v2.6-flash | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 13 | Claude Sonnet 5.5 | Anthropic | 67.2% | Maxeffort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 14 | Qwen3.8 Max (0803) | Qwen (Alibaba) | 67.1% | Defaultvals id alibaba/qwen3.8-max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 15 | Grok 4.6 | SpaceXAI (formerly xAI) | 66.8% | Highreasoning_effort=high | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 16 | GPT-5.6 Sol | OpenAI | 66.5% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 17 | Claude Sonnet 5 | Anthropic | 66.0% | Maxeffort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 18 | Grok 4.7 | SpaceXAI (formerly xAI) | 65.6% | Extra highreasoning_effort=xhigh | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 19 | Gemini 3.8 Flash | Google (Gemini / DeepMind) | 65.3% | Highreasoning_effort=high | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 20 | DeepSeek V4.1 Flash | DeepSeek | 64.3% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 21 | DeepSeek V4 Pro (Preview, 0423) | DeepSeek | 62.9% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 22 | GPT-6.1 Sol | OpenAI | 59.3% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 23 | GPT-6 Luna | OpenAI | 57.6% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |
| 24 | GPT-6 Sol | OpenAI | 56.6% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | — | 29 Sep 2026 |